A 3D Culture Model for Screening of Cancer Therapeutics1K. Patel, 2B. O’Clair, 1T. O’Callaghan, 2D. Appledorn, & 1D. Trezise1Essen BioScience, Welwyn Garden City, AL7 3AX UK & 2Ann Arbor, Michigan, 48108 USA

www.essenbioscience.com

Effects of Lapatinib in a 3D Co-Culture Approach

Quantitative Pharmacology of 3D Spheroids

Differential Effect of Cycloheximide in 2D vs 3D Biology

Cell Player ULA Spheroid Assay Workflow

• Growing evidence suggests the need for physiological relevant 3D models thatmimic the in vivo micro environment (e.g. cell polarity, cell interactions,physiological extracellular matrix).

• Here we describe a 96-well kinetic tumor 3D assay for spheroid growth andshrinkage utilizing Ultra Low Attachment (ULA) microtiter plates (Corning®) inconjunction with the IncuCyte ZOOM.

• CellPlayer™ ULA 3D spheroid assay provides a technically simple, economical, andfully kinetic approach to creating 3D spheroids, equivalent to executing a 2D assay.

• The images and time-lapse movies provide important information and validationregarding spheroid morphology and the effects of test samples.

Summary & Impact

Image Processing & Quantification

• IncuCyte Zoom metrics spheroid size (area, confluence) and fluorescence intensityare informative, and can be readily gathered to monitor spheroid growth andshrinkage over time . Tumor shrinkage is an important translational paradigm fordrug treatment of existing solid tumors.

• These assays should prove useful for medium throughput, quantitative screening oftest samples for effects on the growth and shrinkage of micro-tissues.

• The 3D spheroid assay can be applied as a more physiologically relevant in vitro co-culture model to study the tumor micro-environments.

• Cell proliferation can be directly compared in 2D and 3D.

A549 NucLight Red

Cell Seeding

Harvest and seed cells into 96-well ULA plate (Corning 7007)

• 100 l per well, 1 – 5K cells per well• Centrifuge plate (1000 RPM, 5 minutes

at ambient temperature)

Spheroid Formation

Place ULA plate within IncuCyte ZOOM(4x or 10x)

• Image Phase & Fluorescence• Scan every 6 hr (72-96 hr)

0d 0-3d3d

Sample Addition

Prepare treatments at 2X and add to plate containing spheroids

• Aliqout 100µL 2x treatments to wellscontaining spheroids and media

• Return plate to IncuCyte ZOOM

Spheroid Growth & Shrinkage

Monitor and quantify spheroid parameters over time

• Size (mm2, confluence) & fluorescence intensity metrics

• Real time, automated analysis

3-12d

SSP OX ControlCHX

2D

Cel

l Pro

lifer

atio

n(c

on

flu

ence

)3

D C

ell P

rolif

erat

ion

(Flu

ore

scen

ce c

on

flu

ence

)

-9 -8 -7 -6 -5 -4

0

30

60

90

3D 2D

Log [Cycloheximide] (M)

Inh

ibitio

n o

f g

ro

wth

(%

)

0 50 100 150 200 250 3008

12

16

204.6 nM

14 nM

41 nM

0.12 M

0.37 M

1.1 M

3.3 M

10 M

0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 20 40 60 80 1000

10

20

30

40

50

4.6 nM

14 nM

41 nM

0.12 M

0.37 M

1.1 M

3.3 M

10 M

0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

-10 -9 -8 -7 -6 -50

1000

2000

3000

4000

Log [Staurosporine] (M)

FL

U C

on

flu

en

ce (

AU

C 0

- 2

58h

)

-8 -7 -6 -5 -40

1000

2000

3000

4000

Log [Oxaliplatin] (M)

FL

U C

on

flu

en

ce (

AU

C 0

- 2

58h

)

-9 -8 -7 -6 -5 -40

1000

2000

3000

4000

Log [Cycloheximide] (M)

FL

U C

on

flu

en

ce (

AU

C 0

- 2

58h

)

pIC50 = 7.6Slope = 0.7

pIC50 = 6.1Slope = 2.7

pIC50 = 5.6 Slope = 1.1

Staurosporine Oxaliplatin Cycloheximide A549 NucLight Red Cells were seededin a round bottom ULA plate (5Kcells/well) and allowed to formspheroids over 3 days. Cells weretreated with decreasingconcentrations of either Staurosporine(SSP), Oxaliplatin (OX) orCycloheximide (CHX). Kinetic responsewas plotted as fluorescenceconfluence and EC50 values werecalculated using Area Under the Curve(AUC). The dotted line on theconcentration response curvesrepresents no growth of the spheroid.Note: known cytotoxic compounds,SSP and OX, cause the spheroid toshrink to a size smaller than its startingsize while this is not observed with thecytostatic compound CHX.Quantification was validated byinspection and visualization of images.

A549 NucLight Red Cells were seeded in both a standard flat bottom 96-well plate (2.5K cells/well) and a round bottom ULA plate (5K cells/well). Cells were treated with decreasingconcentrations of either Staurosporine (SSP), Oxaliplatin (OX) or Cycloheximide (CHX). Kinetic response was plotted as fluorescence confluence and the EC50 value of Cycloheximide wascalculated using Area Under the Curve. Looking at the proliferation time-courses we can see the dramatic differences in the temporal profile of CHX in 2D vs 3D. Assay metrics are verified byvisual inspection of the cells, showing striking differences between 2D and 3D proliferation (noted as fold increase from t=0). At the 90h time point, we see selective inhibition of 2Dproliferation (~ 2-fold decrease in cell confluence compared to control) with little or no effect on the proliferation of the spheroid. EC50 data shows a that CHX is almost 10-fold more potent in2D inhibition of growth.

A. A549 NucLight Red cells seeded at2.5K cells/well and fluorescent size wasanalyzed at 72 hours.• Spheroid size at 72h is highly

homogenous within and across platesB. Shown are HD phase images ofspheroids formed from a range of tumorcell lines at 72 hours.• Images validate quantification (verify

size and fluorescent intensity)• Single and centrally located spheroids

can be inspected for differentmorphologies (tight spheroids,compact and loose aggregates)

A. Image Processing• Phase Contrast & Fluorescent images are

acquired.• Due to technical difficulties associated with

masking 3D images in ULA plates,fluorescently labelled cells are used foranalysis/masking.

B. Quantification• IncuCyte ZOOM can automatically quantify

spheroid size (fluorescence area andconfluence). Volume can be mathematicallyderived.

C. Time-dependent reduction in spheroidfluorescence intensity (brightness) is used foragents that cause little or no change in spheroidsize, however cause a marked reduction inspheroid fluorescence.

Consistent Spheroid Formation with the Ability to Inspect Morphology

t=0 t=30 t=60 t=90

Ve

hic

le C

on

tro

lC

HX

1.1

M

Ve

hic

le C

on

tro

lC

HX

1.1

M

2.8-fold 6-fold 7.7-fold

1.6-fold 2.6-fold 4-fold

0.9-fold 1.2-fold 1.3-fold

No change 1.2-fold 1.2-fold

A549 (5K, 72h)

~ 550 m

U-87 MG ( 1K, 72h)

~ 400 m

MCF-7 (5K, 72h)

~ 450 – 550 m

MDA-MB-231 (1K, 72h)

~ 700 – 750 m

A375-C6 (1K, 72h)

~ 750 m

SKOV-3 (7.5K, 72h)

~ 500 m

A172 (5K, 72h)

~ 410 m

RCC4+ (5K, 72h)

~ 460 m

0 50 100 150 200 2500.0

1.0

2.0

3.0

4.0 0.1% DMSO

1 M SPP

Time (h)

FL

U A

rea (

x 1

05

m2)

Fluorescence Area – Measured

Growth

Shrinkage

0 50 100 150 200 2500

5

10

15

20 0.1% DMSO

1 M SPP

Time (h)

Vo

lum

e (

x 1

07

m3)

Volume – Mathematically Derived

Control

0.12 M Mito C

10 M Mito C

Marked reduction in spheroid fluorescence but

not spheroid size

Contr

ol

M M

ito C

0.12

M

Mito

C

10

0

5

10

15

20

25M

IF (

RC

U)

at

198h

Phase/Red blended image

No mask

Phase/Red blended image

Mask overlay

Size

& f

luo

resc

ence

in

ten

sity

incr

ease

s(c

on

tro

l sp

her

oid

)

Shri

nki

ng

& d

imm

ing

(tre

ated

sp

her

oid

)

Exp

1. P

1

Exp

1. P

2

Exp

2. P

1

Exp

2. P

2

Exp

4. P

1

Exp

5. P

1

Exp

7. P

1

Exp

7. P

2

0

100000

200000

300000

400000

Are

a (

m2)

32 or 96 wells, values shown are %CV

A B

5.0 3.8

6.1 5.7

4.7 7.1 5.94.3

Mean ± SEM (3-6 wells)- - - - - No growth

0 50 100 150 200 250 300

5

10

15

20 46 nM1.4nM4.1 nM12 nM37 nM

0.11 M

0.33 M

1.0 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 0.05 nM

0.14 M0.41 M1.2 M3.7 M11 M

33 M

100 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 4.6 nM14 nM41 nM

0.12 M0.37 M1.1 M

3.3 M

10 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 46 nM1.4nM4.1 nM12 nM37 nM

0.11 M

0.33 M

1.0 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 0.05 nM

0.14 M0.41 M1.2 M3.7 M11 M

33 M

100 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 4.6 nM14 nM41 nM

0.12 M0.37 M1.1 M

3.3 M

10 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 46 nM1.4nM4.1 nM12 nM37 nM

0.11 M

0.33 M

1.0 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 0.05 nM

0.14 M0.41 M1.2 M3.7 M11 M

33 M

100 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

0 50 100 150 200 250 300

5

10

15

20 4.6 nM14 nM41 nM

0.12 M0.37 M1.1 M

3.3 M

10 M0.1% DMSO

Time (h)

Flu

ore

scen

ce C

on

flu

en

ce (

%)

Ve

hic

leC

on

tro

lC

HX

10

M

SPP

1

MO

X1

00

M

t=0 t=96h t=198h t=258h

In order to determine the effect of Lapatinib in a 3D co-culture model, SK-BR-3 (breast adenocarcinoma) and CCD-1068SK (mammary derived stromal cells) were first seeded in mono-culture(10k/well) or co-culture (10k/cell line/well) to determine their ability to form spheroids. Images show that CCD-1068SK cells are required for SK-BR-3 to form a compact spheroid (SK-BR-3 cells formloose aggregates in the presence of 5% Matrigel, image not shown). Co-cultures were then seeded and allowed to form spheroids prior to treatment with 3-fold decreasing concentrations of Lapatinib,a tyrosine kinase inhibitor. Fluorescence confluence metrics were used to determine spheroid size, and kinetic data was used to determine Area Under the Curve (AUC). The IC50 value usingthe 3D co-culture method (0.013 µM), is similar to the previously published IC50 value of 0.037±0.031 µM (Konecny, Cancer Res 2006). Interestingly, in a 2D co-culture model, CCD-1068SKfibroblasts have a rescue effect on SK-BR-3 cells from Lapatinib (IC50 = 1.52 µM). Although the 3D co-culture model IC50 did not shift from previously published mono-culture data, imagessuggest that fibroblasts play an important role in tumor formation in the more physiologically relevant 3D in vitro culture model.

CCD-1068SK SK-BR-3 NucLight Red Co-Culture

Spheroid Formation(72 hr)

Arrows indicate re-feed

A B C

Growth

Shrinkage